Wavenumber-Domain Acoustic Reflection Coefficients

• Wavenumber-domain acoustic reflection coefficients are operators defining the amplitude and phase changes of incident acoustic waves via Fourier-transformed pressure fields.
• They generalize traditional models by capturing both specular and non-specular scattering with systematic multiple-scattering corrections applicable to complex materials.
• This framework integrates analytical derivations, numerical estimation, and operator inversion to enhance simulation accuracy in boundary element modeling and material characterization.

Wavenumber-domain acoustic reflection coefficients characterize the amplitude and phase change of acoustic waves upon reflection from boundaries, as functions of incident and reflected wavenumber vectors. These coefficients generalize traditional angle- or frequency-domain reflection models to capture both direction-dependent and non-specular (multi-directional) scattering phenomena, providing a macroscopic operator framework applicable to arbitrary boundary geometries and complex materials, including perforated plates, random particulate media, and multi-species composites. Wavenumber-domain methods offer systematic procedures to derive, estimate, and implement reflection operators in numerical solvers, with the ability to incorporate measured data, analytic models, or high-order multiple scattering corrections.

1. Mathematical Framework and Operator Definition

The wavenumber-domain reflection operator is defined for a planar boundary $\Gamma : z=0$ in a homogeneous medium $(\rho, c)$, relating the spatial Fourier transforms $\hat{p}_i$ and $\hat{p}_r$ of the incident and reflected pressure fields:

$$\hat p_{\rm r}(\mathbf{k}_{xy}) = \iint_{\mathcal{D}_k} C_r(\mathbf{k}_{xy}, \mathbf{k}'_{xy}) \, \hat{p}_i(\mathbf{k}'_{xy}) \, d\mathbf{k}'_{xy}$$

where $$\mathcal{D}_k = \{\mathbf{k}_{xy} : |\mathbf{k}_{xy}|^2 \leq k^2\}$$ is the domain of propagating wavenumbers for given frequency $\omega$ and $k = \omega/c$. $C_r$ is the wavenumber-domain reflection kernel (operator), encoding both specular and scattering effects. In cases where only specular reflection is significant, the operator is diagonal and reduces to a scalar coefficient $$R(\mathbf{k}_{xy}) = C_r(\mathbf{k}_{xy}, \mathbf{k}_{xy})$$ representing the amplitude ratio for each incident direction (Hoshika et al., 12 Jan 2026).

The framework generalizes to non-diagonal $(\rho, c)$0, accounting for direction-changing scattering, diffraction, and material inhomogeneity. The operator norm satisfies energy conservation and causality: $(\rho, c)$1.

2. Derivation for Specific Material Configurations

Closed-form expressions for wavenumber-domain reflection coefficients have been established for several material structures under systematic multiple-scattering and homogenization principles.

Perforated Plates (Low Porosity)

For a doubly-periodic rigid plate with small holes, and under normal incidence, the leading-order coefficients are derived from matched asymptotic expansions and Rayleigh conductivity $(\rho, c)$2 of a single hole:

$(\rho, c)$3

with $(\rho, c)$4 the unit-cell area and $(\rho, c)$5 the effective compliance. The validity regime requires $(\rho, c)$6, $(\rho, c)$7, and $(\rho, c)$8. Higher-order corrections in $(\rho, c)$9 involve lattice-shape constants, but for low porosity, $\hat{p}_i$0 suffices (Laurens et al., 2014).

Random Particulate Half-Space

Ensemble averaging and Wiener–Hopf analysis for a half-space filled with randomly distributed inclusions yield

$\hat{p}_i$1

where $\hat{p}_i$2 is the Wiener–Hopf matrix encoding effective wavenumbers (roots $\hat{p}_i$3) and $\hat{p}_i$4 involves T-matrix elements of individual inclusions (Gower et al., 2019). In dilute-monopole cases, factorized expressions and residue sums reduce the computation:

$\hat{p}_i$5

A plausible implication is that inclusion of multiple wavenumber modes is essential as volume fraction $\hat{p}_i$6 or $\hat{p}_i$7 increases, with reflection accuracy depending on the number of superposed modes.

Multi-Species Materials

For materials with multiple inclusion types, multipole expansion and QCA yield the effective wavenumber

$\hat{p}_i$8

and the reflection coefficient up to second order:

$\hat{p}_i$9

where $\hat{p}_r$0 and $\hat{p}_r$1 are ensemble-averaged far-field amplitudes and multiple-scattering kernels (Gower et al., 2017). The cross terms in $\hat{p}_r$2 capture essential inter-species scattering not present in naïve self-consistent models.

Background-Discontinuous Random Particulate

For a random suspension in a half-space with a background mismatch, QCA and its extension (X-QCA) yield

$\hat{p}_r$3

with explicit coefficients $\hat{p}_r$4, $\hat{p}_r$5, $\hat{p}_r$6 (based on density and wavenumber contrasts) and a multipole-summed backscattering operator $\hat{p}_r$7 (Piva et al., 2024).

3. Estimation, Measurement, and Numerical Implementation

The estimation of $\hat{p}_r$8 or $\hat{p}_r$9 from empirical data involves illuminating the boundary with multiple incident fields, recording reflected pressure distributions, and applying spatial wavenumber transforms. For $$\hat p_{\rm r}(\mathbf{k}_{xy}) = \iint_{\mathcal{D}_k} C_r(\mathbf{k}_{xy}, \mathbf{k}'_{xy}) \, \hat{p}_i(\mathbf{k}'_{xy}) \, d\mathbf{k}'_{xy}$$0 sources and $$\hat p_{\rm r}(\mathbf{k}_{xy}) = \iint_{\mathcal{D}_k} C_r(\mathbf{k}_{xy}, \mathbf{k}'_{xy}) \, \hat{p}_i(\mathbf{k}'_{xy}) \, d\mathbf{k}'_{xy}$$1 wavenumber bins, the data matrices $$\hat p_{\rm r}(\mathbf{k}_{xy}) = \iint_{\mathcal{D}_k} C_r(\mathbf{k}_{xy}, \mathbf{k}'_{xy}) \, \hat{p}_i(\mathbf{k}'_{xy}) \, d\mathbf{k}'_{xy}$$2 and $$\hat p_{\rm r}(\mathbf{k}_{xy}) = \iint_{\mathcal{D}_k} C_r(\mathbf{k}_{xy}, \mathbf{k}'_{xy}) \, \hat{p}_i(\mathbf{k}'_{xy}) \, d\mathbf{k}'_{xy}$$3 suffice:

$$\hat p_{\rm r}(\mathbf{k}_{xy}) = \iint_{\mathcal{D}_k} C_r(\mathbf{k}_{xy}, \mathbf{k}'_{xy}) \, \hat{p}_i(\mathbf{k}'_{xy}) \, d\mathbf{k}'_{xy}$$4

Discretization and regularization are needed when $$\hat p_{\rm r}(\mathbf{k}_{xy}) = \iint_{\mathcal{D}_k} C_r(\mathbf{k}_{xy}, \mathbf{k}'_{xy}) \, \hat{p}_i(\mathbf{k}'_{xy}) \, d\mathbf{k}'_{xy}$$5 or noise is present. This nonparametric approach allows direct ingestion of measured or simulated data, bypassing detailed material microstructure modeling (Hoshika et al., 12 Jan 2026).

4. Transformation to Admittance Operators and Integration into BEM

To employ wavenumber-domain reflection coefficients in Boundary Element Methods (BEM), $$\hat p_{\rm r}(\mathbf{k}_{xy}) = \iint_{\mathcal{D}_k} C_r(\mathbf{k}_{xy}, \mathbf{k}'_{xy}) \, \hat{p}_i(\mathbf{k}'_{xy}) \, d\mathbf{k}'_{xy}$$6 is converted to an admittance operator $$\hat p_{\rm r}(\mathbf{k}_{xy}) = \iint_{\mathcal{D}_k} C_r(\mathbf{k}_{xy}, \mathbf{k}'_{xy}) \, \hat{p}_i(\mathbf{k}'_{xy}) \, d\mathbf{k}'_{xy}$$7 relating pressure and normal velocity spectral components:

$$\hat p_{\rm r}(\mathbf{k}_{xy}) = \iint_{\mathcal{D}_k} C_r(\mathbf{k}_{xy}, \mathbf{k}'_{xy}) \, \hat{p}_i(\mathbf{k}'_{xy}) \, d\mathbf{k}'_{xy}$$8

where $$\hat p_{\rm r}(\mathbf{k}_{xy}) = \iint_{\mathcal{D}_k} C_r(\mathbf{k}_{xy}, \mathbf{k}'_{xy}) \, \hat{p}_i(\mathbf{k}'_{xy}) \, d\mathbf{k}'_{xy}$$9 encodes the propagation normal component. Fourier transforms and whitening procedures ensure robust inversion and accurate imposition of nonlocal, direction-sensitive boundary conditions (Hoshika et al., 12 Jan 2026). Algorithmic complexity scales as $$\mathcal{D}_k = \{\mathbf{k}_{xy} : |\mathbf{k}_{xy}|^2 \leq k^2\}$$0 for operator inversion, but subsequent application in BEM is efficient and bypasses volumetric discretization.

5. Physical Interpretation and Related Multiple Scattering Theories

Wavenumber-domain reflection coefficients synthesize the effects of multiple scattering, boundary-induced mode conversion, and effective media principles:

• In the ensemble-averaged regime, the reflected field is the superposition of plane waves with effective wavenumbers determined via Wiener-Hopf or QCA-based dispersion relations.
• Multiple-scattering corrections, essential at finite inclusion concentrations or with heterogeneous species, appear explicitly in higher-order terms of $$\mathcal{D}_k = \{\mathbf{k}_{xy} : |\mathbf{k}_{xy}|^2 \leq k^2\}$$1 and $$\mathcal{D}_k = \{\mathbf{k}_{xy} : |\mathbf{k}_{xy}|^2 \leq k^2\}$$2 (Gower et al., 2019, Gower et al., 2017).
• Naïve single-wave approximations can introduce significant error; full operator-based or residue-sum approaches are needed for quantitative accuracy as non-dilute or high-frequency regimes are approached.

A plausible implication is that for realistic materials, the complete reflection operator must be estimated or computed including all relevant modes, especially near interfaces or in proximity to boundary layers.

6. Applications, Limitations, and Extensions

Key applications of wavenumber-domain acoustic reflection coefficients include:

• Architectural acoustics: modeling wall and ceiling reflections with directivity-sensitive boundary operators (Hoshika et al., 12 Jan 2026).
• Material characterization: extracting directional reflection fingerprints of panels, metasurfaces, and geological interfaces.
• Noise control: simulating complex scatterers (gratings, diffusers) without meshing fine geometrical details.
• Computer graphics and virtual environments: efficient sound propagation with angle-dependent boundary effects.

Limitations arise primarily from assumptions of linearity, stationarity, and far-field plane-wave decomposition; evanescent and near-field effects are not captured in operator form. For very high angular resolution, computational cost of operator inversion and sampling ($$\mathcal{D}_k = \{\mathbf{k}_{xy} : |\mathbf{k}_{xy}|^2 \leq k^2\}$$3) becomes substantial. Extended approximations such as X-QCA are under active development for structured backgrounds and non-ideal particle distributions (Piva et al., 2024).

7. Comparison with Classical and Contemporary Models

Wavenumber-domain approaches generalize and unify prior normal-incidence or frequency-domain reflection models. In the dilute, low-frequency regime, classical effective density and modulus limits are recovered. For perforated geometries, the reflection formula reduces to known Rayleigh conductivity-based expressions (Laurens et al., 2014). In multi-species and high-concentration cases, the operator formalism demonstrates the necessity of including cross terms and higher-order corrections ignored in “self-consistent” or Waterman–Truell models (Gower et al., 2017).

Comparative numerical studies reveal high accuracy (cosine similarity $$\mathcal{D}_k = \{\mathbf{k}_{xy} : |\mathbf{k}_{xy}|^2 \leq k^2\}$$4, MSE $$\mathcal{D}_k = \{\mathbf{k}_{xy} : |\mathbf{k}_{xy}|^2 \leq k^2\}$$5) for wavenumber-domain BEM against full-geometry modeling, validating the operator approach for complex boundaries and realistic materials (Hoshika et al., 12 Jan 2026).

In summary, wavenumber-domain acoustic reflection coefficients provide a comprehensive macroscopic parametrization of acoustic boundary behavior, incorporating effects of material microstructure, multiple scattering, and angular dependence via measurable or computable operators. These frameworks support advanced simulation, modeling, and characterization tasks in a wide variety of acoustic contexts.